Owing to the bilinear nature of robust performance conditions, it remains a challenge to effectively design a controller for parametric systems. To overcome this difficulty, we establish a metaheuristic-based design framework in this paper. This framework includes a simple initialization method, detailed search flows, and the associated objective functions for each step. In addition, this method can individually and easily shape the gain characteristics of closed-loop transfer functions, thus lowering the hurdle of control design for complex and uncertain systems. The whole design procedure is validated and illustrated through its application to a drivetrain bench. Numerous trials show that on average a success rate of 70% is achieved in the search for the controller.